FOSDEM fun

Just like the last 2 years, I joined the FOSDEM fun. Just like every year there were lots of awesome beers, chocolates, waffles and of course conference full of awesome hackers from all over the world & their awesome talks. In short, the same old 'awesome' experience. Also I finally got to meet Jens George (phako) in person. Another thing that went it a totally unusual (and unexpected) direction was my home directory. Here is how it went:

Early Sunday morning (8:00) my alarm goes off to wake me up in time for my 10:15 am talk. I manage to get up after half an hour of snoozing the alarm, get ready (including getting a shower out of almost non-functional shower), back-up my slides on USB stick and manage to get to the conference room 10 minutes before the talk along with my room-mates, Marc-Andre and Juerg. Failed to get anything to eat or even a cup of coffee before that.

I open-up the lid of my laptop to find out that it won't be able to wake-up from suspend. No problems, happens all the time even since I moved to Ubuntu so I apply the usual solution: Forced reboot. When the system boots I get into another issues that I had been having ever since I moved to Ubuntu: Soon after I provide the password for my encrypted partition (home and swap), I am told that home partition could not be mounted. No problems, I again apply the usual solution: Hit escape key to get to a root shell, where I do this:


# cryptsetup remove home
# cryptsetup create home /dev/sda6
Enter passphrase:
# mount /home


Normally, at this point I just hit CTRL+d and normal boot sequence resumes and everyone on the laptop lives happily ever after. However, this time thats not what happens. `mount` says:


mount: wrong fs type, bad option, bad superblock on /dev/mapper/home,
missing codepage or helper program, or other error
In some cases useful info is found in syslog - try
dmesg | tail or so


This has also happened before but usually this means I mistyped the password and I just have to repeat the above procedure. However, after trying several times, I keep getting the same message. I give-up and lend Marc-Andre's laptop for the slides show. After the presentation, I tried everything and asked everyone I knew who might have any clues but no help came.

Since I get the same questions each time I tell this story to anyone, I thought I append this blog entry by an FAQ:

Q: Why on earth are you encrypting your partitions:
A: Nokia security policy.

Q: What makes you so sure that you haven't just forgotten the password.
A: Because I keep the exact same password for both the partitions and I am able to successfully decrypt one of them.

Q: Have you tried looking at `dmesg | tail`?
A: Yes, nothing in there.

Q: Have you tried fsck?
A: Yes, that can't make any sense out of the "unencrypted" block device either. Because of this, I suspect something is wrong on the actual (encrypted) device, which fsck can not help with.

Syndicated 2010-02-09 11:21:00 (Updated 2010-02-09 11:21:40) from zeenix

GUPnP AV & GUPnP Vala releases

GUPnP AV 0.5.4 released

Changes since 0.5.2:

- Gracefully handle empty or no 'res' node.
- Function to get the list of descriptors from DIDL-Lite objects.
- More complete comparison for LPCM mime types.
- Fix incorrect type conversion of DLNA flags.
- DLNA flags should not default to a specific DLNA version.
- New APIs to deal with contributor-related properties in DIDL-Lite objects.
- Remove redundant construction methods.
- Add forgotten header to the meta-header file.
- Add and fix gobject-introspection annotations.
- Lots of documentation fixes.
- Many minor non-functional fixes/improvements.

Bug fixes in this release:

1935 - Incorrect conversion while parsing primary DLNA Flags
1934 - DLNA Flags should not default to a specific DLNA version.
1933 - Content type matching fails when additional parameters exist in LPCM
       mime-type
1814 - Add "artists" property to GUPnPDIDLLiteObject

All contributors to this release:

Zeeshan Ali (Khattak) <zeeshanak@gnome.org>
Yakup Akbay <yakbay@ubicom.com>
Zachary Goldberg <zach@zachgoldberg.com>

Download source tarballs from here

GUPnP Vala bindings 0.6.4 released!

Changes since 0.6.2:

- Require and adapt to gupnp-av 0.5.3.

All contributors:

Zeeshan Ali (Khattak) <zeeshanak@gnome.org>

Download source tarballs from here

Syndicated 2010-02-04 14:41:00 (Updated 2010-02-04 14:42:23) from zeenix

Me too

I'm even presenting there.

Syndicated 2010-01-28 11:06:00 (Updated 2010-01-28 11:06:21) from zeenix

Rygel 0.4.10 (A Human Reaction)

Rygel 0.4.10 (A Human Reaction) is out!

Another release in the stable 0.4.x series. Changes since 0.4.8:

- Restart application on SIGHUP.
- Fix cancellation of HTTP streams and in turn a critical leak.
- Correct condition for byte-seeking.
- Correctly handle boolean commandline options.
- Enable check for Vala if any stamp file is absent.
- Fix condition for partial content transfer.
- Correct linker argument order. This should fix the build issue on Gentoo.
- Work-around for an automake bug.
- Limit search slice to result size.
- Fix a spelling mistake in error message.
- Tell valac to put debugging symbols to generated code so gdb can refer to
  lines in the Vala source code.

Bug fixes in this release:

601395 - restart on SIGHUP

All contributors to this release:

Zeeshan Ali (Khattak) <zeeshanak@gnome.org>
Jens Georg <mail@jensge.org>

Download source tarball here.

Syndicated 2010-01-24 20:08:00 (Updated 2010-01-24 20:09:33) from zeenix

Rygel 0.4.8 (Till the Blood Runs Clear) is out!

Here goes the release log:

A bug-fix release in stable 0.4.x series to fix seeking in general and playback
for clients that always seek (Sony PS3).

Dependency-related changes:

- Require valac >= 0.7.10.

All contributors to this release:

Zeeshan Ali (Khattak) <zeeshanak@gnome.org>

Download source tarball from here.

Syndicated 2009-12-22 15:29:00 (Updated 2009-12-22 15:29:31) from zeenix